Poster Z18 - Building an anatomically detailed subthalamic nucleus neuron model through optimization

Introduction: Deep brain stimulation (DBS) targeting the subthalamic nucleus (STN) is an effective therapy for Parkinson's disease; however, the precise mechanisms underlying the therapy remain unclear. Detailed multi-compartment computational models of STN neurons are often used to study how DBS electric fields modulate the neurons. However, existing STN neuron models possess certain limitations in their biophysical accuracy. Therefore, our study aimed to enhance a detailed STN neuron model originally created by Gillies & Willshaw [2006].

Materials and

Methods: As shown in Figure 1A, Our design objectives included incorporating an axon into the neuron structure, updating ion channel distributions based on the experimental literature, and refining biophysical parameters to mirror established electrophysiological characteristics of rodent STN neurons. We found that the addition of an axon to the STN neuron model yielded significant alterations in firing behavior. Therefore, we employed a genetic algorithm to fine-tune the biophysical parameters of the model. The optimization process involved systematically adjusting parameters to achieve a close match with experimental observations of STN neuron electrophysiology.

Results, Conclusions, and Discussions: The optimized model exhibited spontaneous firing, action potential shape, hyperpolarization response, and frequency-current curve that aligned well with experimental recordings (Figure 1B and 1C). We also assessed the general compatibility of the updated biophysics by applying them to multiple distinct STN neuron morphologies derived from 3D anatomical reconstructions. Although morphologies impacted firing behavior to some extent, the updated biophysics reliably preserved key electrophysiological features, highlighting the model's versatility. It also indicates the further need of examining the relationship between morphological features and electrophysiological characteristics. In conclusion, our updated STN neuron model offers an improved representation of STN neuron characteristics. We propose that this tool will be relevant for use in the analysis of STN firing behavior during intrinsic synaptic influences, or extrinsic electrical stimulation from intracellular or extracellular sources. The addition of the axon and the model’s versatility also allows the creation of a population of STN neurons with various morphologies which opens up the possibility of capturing essential aspects of real-world neuron population behavior that were previously unaccounted for. The optimization algorithm and cost function designed for this project could also pave the way for broader applications in computational neuroscience.

Acknowledgements (Optional):